TNP #5 - Teardown & Analysis of Damaged Anritsu K250 40GHz & SHF 123B 65GHz Bias-T Blocks

Поделиться
HTML-код
  • Опубликовано: 29 ноя 2024

Комментарии • 58

  • @mikeselectricstuff
    @mikeselectricstuff 3 года назад +24

    They should put a polyfuse on the DC input

    • @Thesignalpath
      @Thesignalpath  3 года назад +13

      That would be a great. To have to throw away a Bias-T because of an over-current incident is incredibly wasteful (and expensive).

    • @dtiydr
      @dtiydr 3 года назад +13

      But then they won't sell any since they won't break.

    • @SeanBZA
      @SeanBZA 3 года назад +4

      I actually have some RF fuse modules, that will enable you to put a 100mA Littlefuse solder in fuse in line with the current path, and which also do provide a whole range of adaptions for different connector types, as well as a straight through option with no fusing. Sadly only rated for 1GHz operation.
      A polyfuse might introduce too much resistance in the power input path, and as well the resistance of the units will not be constant with temperature and current, so having the tiny fuse would be a better option as a in line block that you can easily replace, or at least put a new fuse into by simply unscrewing a connector end and slipping in a new one, from the stock that comes with the fuse unit. Would prevent this kind of accident, though of course you will always find those who blow the last fuse and run it without, and blow the block up as well.

    • @wrtlpfmpf
      @wrtlpfmpf 3 года назад +3

      Actually I used to work at a company making Bias-T for their own internal use. The solution there was a small circuit which would trip on overcurrent like a polyfuse but cheaper.

    • @Vintaginside
      @Vintaginside 3 года назад

      Maybe a clamping device at the DC side?

  • @xDevscom_EE
    @xDevscom_EE 3 года назад +3

    Afternoon here, perfect time to ponder some GHz mmWave things...

    • @jcims
      @jcims 3 года назад

      i've been tinkering with radio stuff for 10 years now, how is the first time i've seen mmWave. love it.

  • @琳寒雨
    @琳寒雨 3 года назад

    Very nice! Looking forward to see the analysis of the 65GHz SHF Bias-T!

  • @vincei4252
    @vincei4252 3 года назад +6

    Very cool and informative. I was able to score a 5KW water cooled N-Type dummy load on eBay. I measured it to make sure it wasn't open or shorted and I also got 52 ohms. I called the very helpful manufacturer and asked if there was anything to be concerned about and they said 52 ohms was within spec. I guess 11:30 I'm in good company :)

    • @subramanianr7206
      @subramanianr7206 3 года назад

      His bias T resistor measured 52 ohm and that's an approximate as the instrument errors are there as well.
      But then a dummy load has to be 50 ohm exactly for precise measurements like when you use a VNA.
      For amateur radio purposes anything around 50 ohm is okay.
      We strive to have the SWR less than 1.5 after all.
      De VU2RZA

    • @LightningHelix101
      @LightningHelix101 3 года назад

      VSWR of 1.5 requires +/-25ohms from 50, so 52ohm load is not going to break the bank

    • @subramanianr7206
      @subramanianr7206 3 года назад +2

      @@LightningHelix101 1.5:1 means 75 ohm on the higher side and 33.3 ohm on the lower side.
      25 ohm means it's 2:1.
      For real RF engineering you ought to have 50 ohm exactly; you would break the bank for sure.
      For we amateur radio enthusiasts +/- few ohm is surely not an issue.
      Anything less than 1.5:1 is more than sufficient and you don't have to expect more; I wanted to stress on that. 😀
      De VU2RZA

    • @Rob2
      @Rob2 3 года назад

      He is not measuring the DC resistance of his dummy load here, but the total resistance of the dummy load and the DC path of the Bias T.
      And of course, the cable.
      Those will also have some resistance, and he postulates that it might be about 2 ohms in total.
      To really know that, one would have to measure at least one of them separately.
      When you measured your dummy load, you may have had some extra series resistance as well, although 2 ohms would be a lot when measuring some heavy device like that (you would likely not have such a thin wire, and other series components).

    • @vincei4252
      @vincei4252 3 года назад

      @@Rob2 Hi, yes, I know. My brief gross test was to make sure that the device wasn't broken thru misuse. Then again, I can't imagine the leads on the fluke multimeter I used have a resistance of 2ohms :-) FWIW, once I get the water cooling radiator and pump all setup I'll throw it on the VNA and see what it really looks like at low power. This unit is good to 2GHz which is well beyond the frequencies I'm interested in.

  • @felixhirt9284
    @felixhirt9284 3 года назад +4

    With that spring in there pressing on the Ferrite/iron core material, i am wondering if the inductor really was welded to the RF-Connection.
    after all, it came out without resistance, and there were no marks on the RF-pin, where the inductor should have made connection.
    I think that the spring presses the tip of the inductor on the RF-Pin.
    maybe you can drill one open before unscrewing the ports?
    very interesting, thank you!

  • @funkyironman69
    @funkyironman69 2 года назад +2

    3:50 That might be conductive epoxy? Would be interesting if heating it allowed the screw to be removed.

  • @subramanianr7206
    @subramanianr7206 3 года назад

    Very wonderful presentation.
    Thanks 👍

  • @ovalwingnut
    @ovalwingnut 3 года назад

    Always InTeReStiNg and usually over my head. But that make it challenging :O) Thank U

  • @hardwareful
    @hardwareful 3 года назад +2

    I bought some of the capacitors shown around 7:30 - muRata UWSC (1nF: 935154521410). The ESL of those goes down to < 20 pH

  • @MrJef06
    @MrJef06 3 года назад

    It looks like the inside of the golden screws is threaded. Interesting.

  • @razawarrior
    @razawarrior 3 года назад

    If you want to do an apples to apples comparison, I would suggest the Anritsu V255 bias tee.

  • @hzmeister9596
    @hzmeister9596 3 года назад +9

    woah, just looked up the price on octopart. are these really over $1k each? pretty unbelievable that a passive component can cost that much and be so easily damaged. would it be possible to add some kind of inline connector with a fuse to protect it without affecting the performance?

  • @DaruoshAghajaney
    @DaruoshAghajaney 3 года назад

    That was interesting. Cool.

  • @Rob2
    @Rob2 3 года назад

    I wonder what went wrong with the one that you disassembled?
    Sure when there is too much current through a thin wire it will melt and break, but you unwound the entire inductor and it appeared to be all in one piece!
    (even when it showed signs of overheating)
    Did the heat maybe influence the pressure of the contact against the RF conductor?

    • @jonka1
      @jonka1 3 года назад +1

      The rigid core of the inductor was no longer intact. I think that would mean it no longer was the correct length to ensure a mechanical pressure contact onto the rf line.

    • @Rob2
      @Rob2 3 года назад

      @@jonka1 Yes, it must be something like that.
      I wonder what the soldered covers on the RF end were for, as it looks like this whole thing could be assembled without them.
      Maybe only for inspection, or maybe leftovers from an earlier generation where they actually soldered those two contacts?
      It does not appear they were soldered in this device, everything looks like clean gold plated material...

    • @jonka1
      @jonka1 3 года назад

      @@Rob2 I am guessing here as I'm not well versed on these high frequencies. It might be solderd in order to make a (near) perfect outer shell profile to rf passing through the body. There is a thing called "skin effect" which has greater effect as the frequency increases and the body will be part of the ground for the signal path.
      Maybe someone who knows about this could add a comment.

  • @todayonthebench
    @todayonthebench 3 года назад +2

    Makes me curious to how good of a bias T one could make out of one of these cases. After all, if the connectors are good, one could likely do something that at least works, even if it might be a bit abhorrent. Question is just how one would approach it. After all, a bit of a waste to throw away a good case and connectors.

    • @BruceNitroxpro
      @BruceNitroxpro 3 года назад

      Today on the Bench , Always some way, just a zillion dollar solution, right?

    • @todayonthebench
      @todayonthebench 3 года назад

      @@BruceNitroxpro No not really. A bias T is a relatively simple circuit. And the goal here isn't to try to beat the original, but rather build a useful bias T, not a perfect one.
      It is in my opinion a bit wasteful to throw out a perfectly good case and connectors if one could do something useful out of it. Doesn't even have to be a bias T as the end product.

  • @McTroyd
    @McTroyd 3 года назад +1

    Bean counter: Spend all the money by the end of the fiscal year, or you lose it next year.
    Lab tech: Hold my beer... 🍺😁👍

  • @EEEAmin
    @EEEAmin 3 года назад +4

    its 2:40AM...time to watch some microwave video

  • @asefrro8296
    @asefrro8296 3 года назад +1

    First is something like Coilcraft BCL Series conical inductor.

    • @bakagaijin7452
      @bakagaijin7452 3 года назад +1

      Saw a couple of QSFP 40g modules with the same conical inductor bias tee inside. Probably its an industry standard thing.

  • @jdlives8992
    @jdlives8992 3 года назад

    Jesh. Why would someone short out those ?

    • @subramanianr7206
      @subramanianr7206 3 года назад +1

      By passing more current inadvertently 🙄😀😂
      De VU2RZA

    • @jdlives8992
      @jdlives8992 3 года назад

      @@subramanianr7206 why

  • @chidon7465
    @chidon7465 3 года назад +1

    Put it in a vice and blast the screws with a heat gun!

  • @joopterwijn
    @joopterwijn 3 года назад

    Sure expected you stray to solder out one of the plugs…. 😂 and,….. yep an other connection problem

  • @ogitakasi3030
    @ogitakasi3030 3 года назад

    Noooice

  • @daveb5041
    @daveb5041 3 года назад

    *How much $ do you have invested in that crazy high GHz test equipment? Also what kind of accent do you have french* ?

  • @gacherumburu9958
    @gacherumburu9958 2 года назад

    👍👍

  • @daveb5041
    @daveb5041 3 года назад

    How much do those things cost ? When I hear expensive I think several hundred? What are/could they used for; cell towers that support 5G's 10's of GHz signals?

    • @Thesignalpath
      @Thesignalpath  3 года назад +4

      They cost about $1K each when new. Normally they are used for lab purposes, testing, instrumentation, etc.

  • @tamaseduard5145
    @tamaseduard5145 3 года назад

    👍👍💖🙏💖👍👍

  • @johnlegros1586
    @johnlegros1586 3 года назад

    classic bad use through too much current by young user !

  • @thijsstavenga4350
    @thijsstavenga4350 3 года назад

    3rd

  • @vincei4252
    @vincei4252 3 года назад

    1st ?